Heavy AI Chatbot Use Has Fallen 31% in the Past Year, Survey Finds

Full text

It's hard to get away from AI these days, with the technology finding its way into our apps, devices, and entertainment — but according to a survey carried out by TechRadar's parent company Future, heavy use of chatbots such as Gemini and ChatGPT has actually fallen over the past year.

The reasons behind the shift are interesting too: users are mentioning concerns over privacy , a preference for human interaction, and worries about becoming too dependent on the tech when it comes to why they're avoiding AI.

While AI is clearly reshaping our world in significant and myriad ways , there's also now a noticeable trend of anti-AI sentiment that's worth paying attention to — whether it's to do with protests against data center expansion or worries about AI slop content .

These results are based on questionnaires filled out by 1,008 respondents, and they make for interesting reading — especially when you compare them to the statistics from the same Future survey conducted last year. Here's what the latest responses tell us.

Privacy worries

There are many reasons why people aren't so sure of AI at the moment. Topping the list for our respondents are concerns over privacy , which were cited by 32% of those who completed our survey — that's actually unchanged from last year, so no matter which chatbot is being used , many of us are wary about what it's doing with our data.

Second in the list, mentioned by 31% of survey participants, is a preference for human interaction (this is a new response option, so we don't have anything to compare it to from last year). While AI is certainly now more than capable of holding a conversation on just about any topic imaginable, it seems we're still seeking out flesh-and-blood relationships .

In third place we've got those who are happy without AI , which accounts for 29% of people surveyed — that's actually a drop of 18% over the previous survey, perhaps a sign that a significant chunk of people have realized they are happy chatting to AI on a daily basis (it's certainly getting more ubiquitous by the day).

There are plenty of people holding out on AI apps like ChatGPT (Image credit: Getty Images / NurPhoto) Next on the list are fears around becoming too dependent on the tech , and concerns over AI misrepresenting a 'voice' or personality (in emails, for example), which account for 26% and 24% of respondents respectively. It seems a lot of us are worried we'll get to the stage where we can't do without AI for even basic tasks — not good when there's an outage — and that our communications will turn into generic, bland AI averageness .

Neither of those two responses were available last year, but the next two came in at exactly the same level as in the previous survey: a lack of awareness of AI's capabilities (19% of respondents) and a lack of interest in what AI could do (17% of respondents). It seems the AI hype hasn't quite reached everyone yet.

Rounding out our respondents' list of barriers to AI usage, we've got technical complexity (15%, down 24%), skepticism over whether AI can be helpful (14%, down 21%), environmental and energy impact concerns (11%, a new category), and being philosophically opposed to AI as a concept 10%, down 24%).

Regular AI usage

When answering the question of how often chatbots like Gemini or ChatGPT are used, 17% of those quizzed said "daily or almost daily", while 14% said several times a day. That's nearly 1 in 3 people using generative AI at least once a day or so, though that's actually fewer people than last year — could the novelty of the tech be wearing off?

The 14% figure for the most enthusiastic users has actually dropped by 31% over the course of the past 12 months, so while usage is exploding in areas such as coding, it seems as though the overall trend is in the opposite direction (perhaps influenced by the increasing amount of AI stuffed into our apps ).

As for the less frequent users, 21% of respondents said they were using AI chatbots once a week or a few times a week, 11% ranked their usage as being a few times each month, and 8% said they used generative AI less than once a month — apparently dipping in and out for occasional chats.

AI chatbots like Claude have faced questions over security and privacy (Image credit: Anthropic) When it comes to complete refuseniks, 30% of the survey participants said they don't use AI chatbots at all, which is a hefty 24% jump from last year. That suggests more and more people are taking against the very idea of AI , or finding it completely useless for day-to-day life and work.

There is a divide in the respondents worth mentioning between those who read Future publications and those who don't: more than 42% of Future readers use generative AI on a daily basis, with 21% using it multiple times a day, which overall is around double the figure for non-readers.

So that's the state of play for consumers and AI at the moment then, according to the people we surveyed. With generative AI advancements showing no signs of slowing down , it's going to be very interesting to see how these figures might have changed again in another 12 months.